Does it really affect the resulting numbers though?
Well, that
really depends on what you want to use those numbers for.
How many people are going to use one of their half dozen characters to create things they aren't remotely interested in and have no desire to play?
With respect, assuming they won't is
just as bad as assuming they will.
The way to know if it impacts the numbers would be to compare numbers with, and without, a filter for active characters, and see if the results differ. This, of course, we cannot do ourselves - D&DB would need to do it for us.
But, the fact that we can't do it doesn't mean we can honestly go ahead without that knowledge, and claim we have an accurate picture.
Especially because, to be honest, we are unlikely to even think of all the relevant possibilities, and the nuances they create.
For example, last I saw D&DB describe it, it measures active characters
in a given time period. Like, active characters this month, this quarter, or this year. I have D&D Beyond characters I built in the past, used for a while
and are now inactive. So, clearly, "never intended to play" is by no means the only option.
Last I recall, the rate of campaign failure is pretty high - a lot of games, start, run for a couple of sessions, and then stop. We should not assume those characters instantly get removed - they may well stay until the user runs up against the limit, and then they'll delete one to make space. Those are characters they had interest in playing, but didn't play long for some reason.